Nvidia stock has become the gravitational center of the AI boom. With a market value above $4 trillion and an order book running into the hundreds of billions of dollars, NVDA now influences not just the semiconductor sector, but the entire equity market. At the same time, fresh export–control drama, a surging Chinese “Nvidia clone,” and persistent “AI bubble” worries are forcing investors to ask: is Nvidia stock still a buy at these levels?
Below is a deep dive into Nvidia’s latest stock price, news, risks, and Wall Street forecasts as of December 6, 2025.
Nvidia stock price and valuation as of December 6, 2025
Nvidia (NASDAQ: NVDA) closed at about $182.41 per share on December 5, 2025, the last trading day before this writing. That implies a market capitalization around $4.4–$4.5 trillion, keeping Nvidia in contention for the title of the world’s most valuable public company. [1]
Key snapshot:
- Closing price (Dec 5, 2025): $182.41
- Market cap: roughly $4.4 trillion
- 52‑week range: about $86.62–$212.19
- Trailing P/E: ~45x
- Dividend yield: effectively 0% (quarterly dividend of $0.01) [2]
After briefly touching $4 trillion in July 2025 and later surpassing $5 trillion in late October, Nvidia’s value has cooled slightly as AI sentiment has become more volatile. [3] Data from several market‑cap trackers shows Nvidia’s valuation down around 7–13% from its recent peak over the last month, but still up almost 30% over the past year. [4]
This combination—slight pullback, still‑huge gains, and a P/E in the mid‑40s—forms the backdrop for the latest debate: is this still reasonable “growth at a high price,” or the late stages of an AI bubble?
The latest Nvidia news investors are watching this week
1. Big fund trims Nvidia, but keeps it as its largest holding
A new filing shows that South Korea’s National Pension Service (NPS)—one of the world’s largest public pension funds—cut its Nvidia stake by 2.4% during Q2, selling about 1.14 million shares. Even after trimming, NPS still owns roughly 46.5 million Nvidia shares, worth about $7.35 billion, representing 0.19% of Nvidia’s shares and a hefty 6.3% of NPS’s total portfolio, making NVDA its single largest holding. [5]
MarketBeat’s summary of analyst ratings in the same article emphasizes how extreme sentiment still is: 54 Wall Street analysts cover the stock; of those, 51 rate it “Buy” or better, with just 2 Holds and 1 Sell, and an average 12‑month price target of $258.65, implying roughly 42% upside from current levels. [6]
So you have a giant, sophisticated investor taking a little profit at the margins, while the analyst community remains overwhelmingly bullish.
2. AI server demand still looks red‑hot
If you’re trying to guess whether Nvidia’s AI party is over, follow the money flowing into the hardware that uses its chips.
Hon Hai Precision Industry (Foxconn), one of Nvidia’s key server manufacturing partners, reported a 26% jump in November revenue to roughly $27 billion, with management guiding to Q4 sales up about 14%. An AI‑focused report on the numbers makes it explicit: much of this growth is tied directly to Nvidia‑based AI servers, and “AI budgets are still open” as orders keep moving. [7]
Even if some software‑side AI projects are struggling to show ROI, the infrastructure build‑out—data centers, racks, networking, power—remains in full swing, which is bullish for Nvidia’s core data‑center GPU business.
3. Geopolitics front and center: SAFE CHIPS, export taxes, and the China question
Nvidia’s biggest risk isn’t just competition; it’s policy.
A new U.S. bill to lock in export curbs.
On December 4, a bipartisan group of U.S. senators introduced the SAFE CHIPS Act, which would block the Trump administration from loosening restrictions on AI chip exports to China, Russia, Iran, and North Korea for 30 months. The bill would require the Commerce Department to deny licenses for AI chips more advanced than what those countries can already buy, and to brief Congress before any future rule changes. [8]
In other words, even if the current administration wants to ease pressure, Congress is trying to bolt the door—directly affecting Nvidia’s ability to sell high‑end GPUs into one of its previously most important markets.
Export taxes and China’s pivot to domestic chips.
Separately, Reuters reports that in August 2025 the Trump administration struck a deal with Nvidia and AMD: they could resume shipping certain AI chips to China if they agreed to pay a 15% fee to the U.S. government. AMD CEO Lisa Su confirmed that AMD has licenses to ship its downgraded MI308 chips under this framework and is prepared to pay the tax. [9]
China, however, has pushed back:
- Beijing issued guidance that state‑funded data centers must use homegrown AI chips, a move that directly undercuts future demand for Nvidia, AMD, and Intel in China. [10]
- Tom’s Hardware, citing Bloomberg, reports that Nvidia CEO Jensen Huang is now uncertain whether China would even allow firms to buy Nvidia’s H200 GPU if export rules are loosened. He also stated Nvidia cannot simply “degrade” chips for China, because customers there won’t accept inferior products. [11]
- Earlier this year, export bans and counter‑measures led to Nvidia’s H20 China‑specific GPU being banned, a $5.5 billion write‑off, and China subsequently ordering tech firms to stop buying certain Nvidia chips, including the RTX Pro 6000D. Huang says Nvidia’s AI GPU market share in China has fallen from about 95% to “essentially zero.” [12]
The near‑term bottom line: Nvidia’s Q4 guidance assumes essentially no China data‑center revenue, and future upside from China depends on both U.S. export policy and Beijing’s willingness to allow American chips back into the most sensitive parts of its AI stack. [13]
4. A rising Chinese rival: Moore Threads’ explosive IPO
Against that backdrop, a Chinese GPU designer with deep Nvidia DNA just had a spectacular debut.
- Moore Threads, founded in 2020 by former Nvidia executive Zhang Jianzhong, designs GPUs for gaming and AI workloads.
- The company’s Shanghai STAR Market IPO raised over $1 billion, and the stock closed its first day up about 425%, giving it a market cap around $40 billion. [14]
- Barron’s notes that Moore Threads’ latest S4000 AI chip still lags Nvidia’s older H100 GPU, and the company faces big domestic rivals like Huawei and Cambricon. Meanwhile Nvidia itself is valued around $4.4–$4.5 trillion—roughly 100x Moore Threads’ size. [15]
This doesn’t threaten Nvidia’s global dominance today, but it underscores two realities:
- China is rapidly building its own GPU ecosystem.
- Even if export rules are relaxed, domestic Chinese chips will be politically favored for critical infrastructure.
Both factors cap how much of Nvidia’s long‑term addressable market will actually be accessible in China.
5. Onshoring AI chips: the first U.S.‑made Blackwell wafer
On the other side of the geopolitical chessboard, Nvidia is leaning hard into U.S.-based manufacturing.
On October 17, 2025, Nvidia and TSMC announced the first Nvidia Blackwell wafer produced at TSMC’s new fab in Phoenix, Arizona, marking Blackwell’s volume production and a milestone in “onshoring” the AI hardware stack. [16]
- TSMC Arizona will manufacture 2nm, 3nm, and 4nm chips, including Blackwell GPUs, for AI and high‑performance computing. [17]
- Jensen Huang called it “the very first time in recent American history that the single most important chip is being manufactured in the United States,” framing it explicitly as part of a broader U.S. reindustrialization push. [18]
For Nvidia shareholders, this helps:
- Reduce supply‑chain risk tied to East Asia.
- Align Nvidia with U.S. industrial and security policy, which may matter as export controls evolve.
Earnings recap: record Q3, massive order book
Nvidia’s fundamentals remain staggering.
In its Q3 fiscal 2026 results (quarter ended October 26, 2025), Nvidia reported: [19]
- Revenue: $57.0 billion
- Up 22% QoQ and 62% YoY
- Data Center revenue: $51.2 billion
- Up 25% QoQ and 66% YoY
- GAAP gross margin: 73.4%
- GAAP EPS: $1.30 (ahead of analyst estimates)
Guidance for Q4 FY26 was equally bold:
- Revenue guidance: $65.0 billion, plus or minus 2%
- Expected gross margin: ~75% (GAAP and non‑GAAP) [20]
The earnings call and commentary added some critical color:
- CFO Colette Kress reiterated that Nvidia has “visibility to a half‑trillion dollars” in Blackwell and Rubin AI chip revenue from the start of 2025 through the end of 2026, and said that number is likely to rise as more deals are signed. [21]
- She also emphasized that hyperscalers like Meta, Microsoft, Amazon, and Google account for roughly half of Nvidia’s long‑term opportunity, as workloads transition from CPU‑heavy to GPU‑accelerated computing. [22]
- The company highlighted multi‑gigawatt infrastructure deals, including OpenAI planning at least 10 gigawatts of Nvidia‑based systems and a deep technology partnership with Anthropic, where Nvidia may invest up to $10 billion as Anthropic spends $30 billion on Nvidia‑powered compute. [23]
Analysts quoted after the call described the results as “blockbuster” and argued that the AI revolution “is nowhere near its peak”, while also warning that power, land, and grid constraints could become bottlenecks by 2026 and beyond. [24]
Nvidia itself, in its 10‑Q, flagged energy infrastructure as a potential constraint on how fast AI data centers can scale, calling out the multi‑year timelines required to upgrade power grids. [25]
Nvidia’s answer to “AI bubble” fears: the full-stack moat and Rubin
The “AI bubble” phrase is everywhere, including in academic reports. A 2025 analysis from MIT highlighted that despite $30–40 billion in enterprise generative‑AI investment, around 95% of organizations still see little or no direct financial return, fueling concerns that the sector could be over‑hyped. [26]
Nvidia’s leadership is aggressively pushing back on that framing.
At a recent UBS Global Technology and AI Conference, CFO Colette Kress argued that what we’re seeing is not a speculative bubble but a structural transition from CPU‑centric to GPU‑accelerated computing. She emphasized: [27]
- Nvidia isn’t just selling a chip; it runs “seven different chips” together across the AI lifecycle (training, fine‑tuning, inference), anchored by the CUDA software ecosystem.
- That full stack, she said, makes Nvidia’s AI platform difficult for ASIC‑based competitors to dislodge, even as specialized chips proliferate.
Kress also confirmed that Nvidia’s next‑generation Vera Rubin architecture has taped out, and the company is “working feverishly” to bring Rubin systems to market in the second half of 2026. [28]
Put simply: Nvidia’s internal view is that we’re still early in a multi‑year transition, and that its advantage is not just in one flagship GPU, but in the tightly‑integrated hardware‑plus‑software stack.
What Wall Street expects from Nvidia stock (2026–2030)
Different aggregators vary slightly, but the broad picture from analysts is remarkably consistent: huge growth, strong buy ratings, and meaningful upside from here—with lots of caveats between the lines.
12‑month price targets
- MarketBeat (54 analysts):
- Consensus rating: Buy (with 5 “Strong Buy,” 46 “Buy,” 2 “Hold,” 1 “Sell”)
- Average 12‑month price target:$258.65, implying about 42% upside from $182.41
- Target range: $205–$352 [29]
- StockAnalysis (39 analysts):
- Consensus rating: Strong Buy
- Average target:$248.64, implying roughly 36% upside
- Median target ~$250; high $352; low $100 [30]
- 24/7 Wall St.:
- Notes a median Street target of about $257.72 (~43% upside)
- Its own model sets a 2025 fair‑value estimate at $233.16 (about 30% upside), based on projected EPS of $2.75 and a 50x P/E. [31]
Different methodologies, same general message: analysts as a group still expect NVDA to rise 30–45% over the next year.
Fundamental growth forecasts
Where things get more extreme is in the fundamental forecasts.
- StockAnalysis’ consensus model projects Nvidia’s revenue rising from about $130.5 billion in FY 2025 to around $217 billion in FY 2026, then to roughly $331 billion in FY 2027—a staggering two‑year stretch of ~66% and ~52% growth. [32]
- EPS is forecast to climb from roughly $2.94 to $4.77 to $7.76 over the same period. [33]
- 24/7 Wall St.’s longer‑term model projects revenue hitting about $265 billion and net income ~$175 billion by 2030, with EPS around $7.24 and a base‑case stock price of $318.42, about 77% above today’s price. Its “high” scenario imagines Nvidia trading around $506 by 2030 (70x earnings), while a “low” scenario around $217 assumes a more modest 30x multiple. [34]
These are forecasts, not facts. But they show what’s baked into the current valuation: the market is implicitly assuming Nvidia can more than double its already enormous revenue base within just a few years.
Key forces that could move Nvidia stock next
Rather than pretending there’s a single neat answer to “buy or sell,” it’s more honest to lay out the main forces pulling NVDA in each direction.
Bullish forces
- AI infrastructure supercycle
Data‑center demand is still surging. Hon Hai’s November numbers, Nvidia’s own record Q3, and multi‑gigawatt deals with OpenAI, Anthropic, and hyperscalers all point toward continued heavy capex on AI hardware. [35] - Full‑stack moat (chips + CUDA + networking)
Nvidia doesn’t just sell GPUs—it sells a platform: GPUs, CPUs, networking, software, and libraries, all optimized around CUDA. That makes switching costly for customers and helps explain why analysts still see multi‑year dominance despite AMD, Google TPUs, and specialized ASICs. [36] - Onshoring and political alignment
Blackwell production at TSMC Arizona is not just symbolic. It strengthens Nvidia’s position in U.S. industrial policy debates and helps reassure hyperscalers and governments that the AI supply chain won’t be hostage to a single geography. [37] - Order book and visibility
The company’s claim of $500 billion in Blackwell and Rubin revenue visibility through 2026, combined with consensus forecasts of revenue crossing $300 billion by 2027, suggests that much of Nvidia’s near‑term growth is already “spoken for” by customers. [38]
Bearish (or at least cautionary) forces
- Valuation and concentration risk
Nvidia’s market cap, which briefly exceeded $5 trillion, is larger than the GDP of most countries, and its weight in the S&P 500 has raised concentration concerns among regulators and asset managers. [39] If AI expectations reset—even from “insanely euphoric” to merely “optimistic”—a lot of air could leave the balloon. - AI bubble / ROI concerns
That MIT finding that 95% of enterprises see little direct return on generative AI so far is a reminder that software‑side monetization is still early and uneven. [40] If boardrooms decide that AI budgets are growing faster than realized value, hardware spending could slow faster than current models assume. - Geopolitical and regulatory risk
- SAFE CHIPS and similar proposals could lock in strict export controls for years, effectively capping Nvidia’s addressable market in China and several other countries. [41]
- China’s move to require homegrown AI chips in state‑funded data centers, and the collapse of Nvidia’s AI GPU share there from ~95% to near zero, shows how quickly policy can erase an entire regional business. [42]
- Competition catching up at the margins
AMD is shipping increasingly credible AI accelerators (MI300, MI308), is willing to pay export taxes to access China, and is aggressively courting hyperscalers. Meanwhile, Google, Amazon, and others are scaling their own in‑house silicon. China’s Moore Threads and other domestic startups are still behind, but they’re scaling fast with strong policy support and capital. [43] - Physical bottlenecks: power, land, and energy grid
Nvidia’s own filings stress that energy capacity and grid build‑out could slow AI data‑center deployments, introducing a physical cap on how fast AI spending can grow. [44]
Is Nvidia stock still a buy?
From a purely narrative standpoint, Nvidia is still the clear center of gravity in the AI hardware universe:
- Dominant market share in AI GPUs
- Record revenue and margins
- A gigantic pipeline of orders
- Deep relationships with every major hyperscaler
Analysts, on average, think NVDA can rise another 30–45% over the next year, and longer‑term models see the potential for 70%+ upside by 2030—if revenue and earnings follow the very aggressive trajectories currently penciled in. [45]
But the risks are equally outsized:
- A policy misstep in Washington or Beijing could erase tens of billions in expected revenue.
- A broad AI sentiment reversal could compress Nvidia’s P/E from the mid‑40s toward a more “normal” mega‑cap multiple, even if fundamentals stay strong.
- Physical limits—power grids, cooling, land—may slow the pace at which Nvidia’s half‑trillion‑dollar order book actually turns into reported revenue. [46]
Whether Nvidia is “still a buy” comes down to how you weigh those forces:
- Growth‑oriented investors who believe in a multi‑decade AI productivity wave and Nvidia’s full‑stack moat may view the current pullback from its highs as a consolidation phase in a much longer uptrend.
- More cautious or value‑oriented investors may conclude that a $4‑plus trillion, 45x‑earnings chip company at the center of a potential bubble simply doesn’t leave enough margin of safety, no matter how dazzling the story.
Either way, the key is to treat Nvidia not as a sure thing but as a high‑beta, high‑impact AI infrastructure bet, and size positions accordingly.
Important: This article is for information and analysis only and does not constitute investment advice. Always consider your own financial situation and risk tolerance, and consult a qualified advisor before making investment decisions.
What to watch next for NVDA
Going forward, Nvidia shareholders and traders will be watching:
- Q4 FY26 earnings and updated guidance, especially any changes to data‑center growth expectations and commentary on China. [47]
- The progress (or failure) of the SAFE CHIPS Act and related export‑control legislation in the U.S. [48]
- China’s AI chip policy, including further guidance on domestic‑chip requirements and enforcement against cloud workarounds. [49]
- Updates on Blackwell and Vera Rubin, including production ramps at TSMC Arizona and any early performance or adoption data. [50]
- Evidence—positive or negative—on enterprise AI ROI, which will eventually determine whether current AI capex levels are sustainable. [51]
References
1. companiesmarketcap.com, 2. www.marketbeat.com, 3. www.reuters.com, 4. public.com, 5. www.marketbeat.com, 6. www.marketbeat.com, 7. completeaitraining.com, 8. www.reuters.com, 9. www.reuters.com, 10. www.reuters.com, 11. www.tomshardware.com, 12. www.tomshardware.com, 13. www.businessinsider.com, 14. www.barrons.com, 15. www.barrons.com, 16. blogs.nvidia.com, 17. blogs.nvidia.com, 18. blogs.nvidia.com, 19. investor.nvidia.com, 20. investor.nvidia.com, 21. www.businessinsider.com, 22. www.businessinsider.com, 23. investor.nvidia.com, 24. www.businessinsider.com, 25. www.businessinsider.com, 26. en.wikipedia.org, 27. wccftech.com, 28. wccftech.com, 29. www.marketbeat.com, 30. stockanalysis.com, 31. 247wallst.com, 32. stockanalysis.com, 33. stockanalysis.com, 34. 247wallst.com, 35. investor.nvidia.com, 36. wccftech.com, 37. blogs.nvidia.com, 38. www.businessinsider.com, 39. en.wikipedia.org, 40. en.wikipedia.org, 41. www.reuters.com, 42. www.tomshardware.com, 43. www.reuters.com, 44. www.businessinsider.com, 45. www.marketbeat.com, 46. www.businessinsider.com, 47. investor.nvidia.com, 48. www.reuters.com, 49. www.tomshardware.com, 50. blogs.nvidia.com, 51. en.wikipedia.org
